Chapter 7 - MARISOL’S FILE
Marisol requested her complete employment file seven years after the case.
Why then?
She was applying for an executive role at a national benefits organization.
Background screening found the old termination record in a vendor database despite the settlement correction.
There.
Systems remember badly.
Her official Moretti personnel file had been corrected.
But a third-party screening vendor retained:
terminated for suspected fraud.
Wrong.
She called Nina.
Then me?
No.
Good.
Nina handled.
The vendor had received data from an old Moretti security feed before reforms.
It had never been updated.
There.
Data persistence.
The correction required:
formal dispute,
source update,
vendor deletion,
written certification.
Good.
No phone call from Nicholas.
I learned only because Moretti Holdings had to correct the source system.
Then the company audited all historical termination feeds.
Forty-seven former employees had records requiring review.
Most accurate.
Six outdated.
three materially misleading.
one duplicate identity.
Good.
Not only Marisol.
Then policy:
adverse employment data sent to vendors must include correction/update process.
retention limits.
annual audit.
employee notice where law requires.
Good.
A personal injustice exposed system design.
But reform applied broadly.
Then Marisol got the job.
Without my recommendation.
Good.
National Workers Support Network.
Director of operations.
She moved to Cambridge.
Isabella helped pack.
Mateo complained.
Ordinary.
Then one day Marisol asked to meet me.
Not legal office.
Coffee.
First time without counsel.
The litigation had been closed for years.
We sat.
She said:
“I almost sent my daughter to you because I thought power was safer than process.”
There.
I looked at her.
“You did send her.”
“Yes.”
Then:
“I hate that.”
Good.
No saint mother.
She had made a dangerous contingency plan.
It worked.
That did not make it wise.
Then:
“I should have given Elena better instructions. I should have had a lawyer copy. I should not have put that on Isabella.”
There.
I nodded.
Then:
“You were under pressure.”
“That explains it.”
She said.
“It doesn’t make an eight-year-old the right messenger.”
There.
Good.
We could both admit harm without ranking.
Then I said:
“I’m glad she found me.”
Marisol nodded.
“So am I.”
Both.
Then:
“That doesn’t mean I’d do it again.”
Good.
We drank coffee.
Then she asked:
“Do you regret prison?”
Dangerous question.
“I regret the conduct that led there.”
Specific.
Then:
“Did prison help?”
“Some.”
Good.
No glorification.
Then:
“Would you have changed without it?”
“I don’t know.”
Honest.
Then Marisol said:
“Gabriel wrote me.”
My body stiffened.
“He’s out?”
“Yes. Supervised release.”
There.
Gabriel had served his sentence.
He asked Marisol for a meeting.
She declined.
Good.
No one owes closure.
Then Gabriel wrote again through counsel only once:
an apology.
Marisol read it.
No response.
Fine.
She asked if I had heard from him.
“No.”
Would I meet him?
“I don’t know.”
There.
Then one year later, I did.
Not because he asked repeatedly.
Because I wanted one answer.
Neutral room.
No friendship restored.
Gabriel looked older.
He said:
“You look well.”
I ignored it.
“Why did you use my name?”
He knew what I meant.
The trust.
The managers.
The loyalty language.
He leaned back.
“Because it worked.”
There.
Same as me.
Then:
“Did you think I wanted it?”
“Yes.”
Hard.
“Why?”
“Because for twenty years, when someone said ‘Nicholas will hear about this,’ you never complained.”
There.
Memory.
He was right.
I had benefited from implied force.
Then:
“I didn’t tell you to stop aid to children.”
“No.”
Good.
Then:
“I thought you wanted loyalty to have consequences.”
There.
Again.
I had.
Then I asked:
“Do you think I betrayed you?”
Gabriel laughed.
“You put me in prison.”
“No.”
“Your lawyers—”
“Evidence did.”
There.
He looked angry.
Then:
“I think you betrayed the rules we actually had.”
Fair.
Which rules?
Not law.
Our private rules.
That was the problem.
Private rules had outranked public ones for too long.
Then Gabriel said:
“I helped you build everything.”
True.
Then:
“And you let outsiders tear it apart.”
There.
I answered:
“They found crimes.”
Silence.
Then:
“So did you.”
Good.
No moral high ground.
We ended.
No handshake.
I never saw him again.
Then Moretti family reaction.
Some cousins still called Gabriel loyal.
Some called him thief.
Fine.
I did not enforce opinion.
One cousin asked if Gabriel could attend a funeral.
Not my decision.
The deceased’s immediate family invited him.
I attended too.
We stood across room.
No confrontation.
Good.
Families contain people who disagree about harm.
Then Isabella’s policy coursework.
She emailed me once.
Subject:
Question for class.
She asked:
“Can a private charity condition voluntary aid on signing a legal release?”
Good question.
I replied:
Ask your professor first. In general, sometimes settlement-linked aid can be lawful if terms are disclosed and not coercive, but context matters. I’m not your lawyer.
Then added:
Rebecca can recommend public sources.
Good.
No rewriting her paper.
She replied:
Annoying but useful.
Fine.
Then at twenty-two, Isabella interned at city housing policy office.
No Moretti connection.
Mateo started community college.
He changed majors twice.
Good.
No burden to redeem family history.
Then Marisol sent me a photo.
Mateo at graduation from EMT training? Maybe too destiny. Better from culinary certificate? Let's keep ordinary: high school graduation.
Isabella beside him.
I looked.
Then deleted the photo after saving? She sent, so can keep. Fine.
I did not print.
No shrine.
Children grow whether powerful men are watching or not.
Marisol’s employment-file correction also exposed one more uncomfortable Moretti practice.
Security vendors had retained old “reliability scores.”
Employees rated:
financial pressure,
family instability,
conflict history,
likelihood of departure.
There.
The same mentality as loyalty files.
Some data came from lawful background checks.
The scoring purpose was the problem.
A worker with debt could become:
higher coercion risk.
A worker with childcare difficulty:
higher flight risk.
Those labels then influenced access to sensitive assignments.
Maybe rational in security theory.
Also dangerous.
Independent review found no evidence the scores had been used for pay or termination broadly.
Still, workers had not understood such profiles existed.
The company ended the program.
Replaced with:
role-specific access controls,
actual conflict disclosures,
specific threat assessments,
employee assistance separated from security.
Good.
Then one manager protested:
“What if someone is vulnerable to bribery?”
Security chief answered:
“Then design systems where one person cannot secretly approve high-risk acts.”
There.
Better.
Reduce reliance on guessing who is weak.
Then an employee did report a bribery attempt later.
He had significant personal debt.
Old score would have marked him risky.
In reality, he reported.
Good.
People are not their vulnerability profile.
Then Marisol’s new organization asked her to build an employee emergency fund.
She called me once for one reason:
“What should I avoid?”
I laughed.
Then stopped.
She was serious.
I answered:
“Do not let legal, HR, debt collection, or management performance decisions control aid.”
Then:
“Independent appeals. Clear eligibility. No loyalty language.”
Good.
She said:
“That’s useful.”
No consulting fee.
Fine.
Then her fund made a different mistake:
too many documents required before emergency rent help.
Applicants dropped out.
They changed provisional-assistance rules.
There.
No system is fixed permanently.
Then one employee demanded aid for an expense outside mission.
Denied.
He accused fund of hypocrisy.
Appeal upheld denial.
Good.
Compassion requires scope.
Then I asked Marisol:
“Do you ever wish you hadn’t overridden those payments?”
She thought.
“Some.”
There.
Good.
Then:
“Do you wish you hadn’t reported?”
“No.”
Two truths.
That was mature.
Then she asked:
“Do you wish you never met us?”
No.
I answered:
“I wish Isabella had not needed to find me.”
There.
Better.
The relationship could matter without romanticizing the danger that created it.
The correction project also showed how difficult deletion can be.
One old vendor refused to purge outdated reliability scores because its contract allowed seven-year retention.
Legal had to negotiate:
correction flags,
restricted use,
eventual deletion,
notice to affected employees.
Good.
You cannot always erase bad data instantly.
But you can stop pretending it is current.
Then one former employee said:
“I don’t want a settlement. I want the record corrected.”
Done.
Another wanted compensation for a job lost after an inaccurate fraud flag.
Evidence supported part of the claim.
Independent administrator paid.
Good.
A third blamed a rejected mortgage on Moretti employment data.
Records showed the lender never received it.
Claim denied.
Specific.
Marisol watched the process from outside.
She said:
“This is what I wanted the first time.”
“What?”
“Someone who could say yes or no without needing Gabriel to approve the answer.”
There.
Independence.
Not universal agreement.
Then I thought about how often I had described loyalty as trust.
They were not the same.
Loyalty asks:
Are you with me?
Trust asks:
Can I let you reach a conclusion I may dislike?
I had been good at the first.
The second took longer.
Marisol later told me the corrected file helped more than the settlement money in one unexpected way.
When she interviewed for the national role, she could answer the termination question plainly.
There had been an allegation.
Charges were dismissed.
The employer corrected the record.
An independent review documented retaliation.
No whispered explanation.
No asking a powerful man to vouch for her.
Paper did the work.
Good.
May you like
That was what correction should do:
reduce dependence on someone else’s credibility.
